Rampage is Here!!! ***ENGLISH VERSION***

[Sansones]

Hello to everybody!
If you are visiting this thread you already know that our lab is testing the unique model of the 3dfx Rampage video card and that very soon we will release all the tests.

But let start introducing the whole family of last 3DFX video cards!

[Sansones]

Some days ago, during an important news about the accelerator Spectre 1000, based on the well known Rampage processor, were announced a series of tests of the famous Voodoo 5 6000.

We finally come in possession of two different versions of this video card: the first one is a Voodoo 5 6000 rev. "A" video card, the second one is a Voodoo 5 6000 rev. "A3" video card.

The rev. "A" is the most “common” one and its differences from the rev. "A3" are that it’s equipped with SD-RAM of 6ns and is made in Mexico.
The rev. "A3" in our possession, on the contrary, was all assembled in the USA and is equipped with SD-RAM of 5.4ns (the fastest memory used on that kind of video card)! Moreover, we've more than a reason to think that the one in our possession is the only well working model in whole the world. Last but not the least, it represented the pre-seller version model of the powerful 3dfx graphic accelerator based on the VSA 100 graphic chip.


3DFX Voodoo 5 6000 rev. "A" with SD-RAM V5-6000 "A": 6ns


3DFX Voodoo 5 6000 rev. "A3" with SD-RAM V5-6000 "A3": 5.4ns


Looking at the four pictures above, made by our more and more famous Emanuele Chiocchio (aka "MAN") in the professional photographic lab of NextHardware, we can admire (I think it’s the better expression we can use) on the first line the rev. "A" and on the second line the rev. "A3" (we can notice a different PCB colour).


These cards, we know, have always had some small working problems, few stability and some other little flaws. This kind of bugs were fixed making on the rev. "A" all projectual modifications made up by 3dfx on the more stable and recent Voodoo 5 6000 "A3".
Who was able to do that ? What the secrets hidden behind these accelerators? What are the scores in the different benchmarks?

To discover the truth, stay tuned on NextHardware.com!



*A special thanks goes to our collaborator Francesco V. aka Amigamerlin and to 3DFXZONE. We invite you to visit 3DFXZONE forum: in which you can find more new pictures about the two video card.

[Sansones]

Rampage has arrived in Italy, the only really working model and the only model produced of that revision (the last one). I don’t think is necessary to say something more!

[Sansones]

3DFX Rampage

The rumour of the existence of two pictures of the Rampage, the graphic accelerator which, in the past, would have had to change the 3DFX economic destiny, published in 3DFXZone.it, is jumping over the whole net.

We can confirm that the pictures we are talking about are absolutely original and that the publisher is a more than reliable font; it’s the developer of 3DFX actually most used drivers: Amigamerlin.

The other news is that NextHardware.com is in possession of the whole technical datas of the Rampage and that those datas will be published before this night.

In some days, we’ll have in our labs a Voodoo 5 6000 A3 card by which we can do very interesting tests, followed by an accurate article about the story of the development of technologies implemented in actual graphic card accelerators: we will surprise you with our further news.

[Sansones]

3DFX Rampage

As we promised, here you are the “official” datas of the 3DFX Rampage video card, when it was, as everyone in the world community was believing, only a project! ...but we know that this is not the truth…

These datas are taken from a Power Point original file come in our possession. That file was made as information for investors.

We have literally been amazed, as the first time in which such characteristics were made, above all, known as a "rumours" (therefore however not confirmable).


Rampage

Next Generation Rendering Engine (VTA)
Next Generation Video Engine
16M tri/sec, 800MP/sec, Single textured.
Programmable Shading
Higher quality filtering
T-buffer based AA
AA With 2 textures and Anisotropic with no performance hit.
AA With 4 textures with no performance hit
Scalable Architecture
Integrated Bridge functionality
Full AGP Support in SLI
Two recursion modes
Before and after W divide
Applicable for bump mapping and procedural textures
Risc-like address op instruction model
Cubic Environment Mapping
Texture Compression
DirectX/DXT ,FXT1 compressed textures
YUV Texture formats
Massive Multitexturing: 8 textures each with different
filtering: point, bilinear, trilinear, anisotropic (up to 128-tap).
Massive Iterators: 8 ARGBSTW iterators: one per texture.
Texture Recursion (Address ops) : recursion in parallel or series on up to 4 textures
Any combination of 4 recursions on the 8 available textures Recursion allows one texture to perturb the values of another texture
Texture Computer (Pixel Shader)
Massive flexible math capability
texture combine unit: (a+b)*c+d feeding...
...color combine unit: (A+B)*C+D where each abcd and ABCD can be:
from any of iterator, texture, previous, registered and constant values.
manipulated independently with: x, -x, 1-x and x-0.5.
Compare based multiplexing:
texture and color: combine unit: (a-b)? c:d and
texture combine unit output summing
R+G+B may feed
color combine unit: .(a+b)*c+d, alpha and/or color channels Useful for intensity ops, dot products.
Register based combiners
Larger color component precision
13 bits per component
Precision maintained through alpha blenders.
Alpha Blender Logic Ops
Quad pixel pipeline
4 pixels-per-clock single textured (per chip)
4/N pixels per clock for N textures (per chip)
Scalable rendering: Up to four chips ganged together. No bridge chip required.
Optimized Geometry Pipeline.
Viewport Transforms and efficient host/graphics IO yield optimum performance
Works with (or without) Sage geometry co-processor
Larger color component precision
13 bits per component
Precision maintained through alpha blenders.
Alpha Blender Logic Ops
Quad pixel pipeline
4 pixels-per-clock single textured (per chip)
4/N pixels per clock for N textures (per chip)
Scalable rendering: Up to four chips ganged together. No bridge chip required.
Optimized Geometry Pipeline.
Viewport Transforms and efficient host/graphics IO yield optimum performance
Works with (or without) Sage geometry co-processor

ecc...

Many other datas are informations and specs for programmers, so we've decided to insert here only most comprehensible datas that can show the capacity and the power of the Rampage.


In the second photos: the heat sink without fan is a prototype used only for lab tests. This system frees the tester to change Rampage processors in a very simple way without soldering or desoldering the chip. This kind of prototype has a value market of about 1200 US Dollars.


News are not ended! We hope we will enjoy you with many other news in the next hours. The purpose of our portal and our job is the information and the communication about NEW technology and 'possible/secret' technology.
All the users that follow us can enjoy therefore of very fresh informations; we will do the possible (and the impossible) to be able to enrich as much as we can such "cake" that seems to be really well stuffed.


*A special thanks goes to our collaborator Francesco V. aka Amigamerlin who is working very hard to let us to know all the most interesting news about the Rampage matter.

[Sansones]

3DFX Spectre 1000: experimental board.

Here we are with other fresh news about the already well known RAMPAGE!

The picture, a world exclusive, shows the heat sink system and the chip aside taken off the Rampage board.
This board is provided with a real socket on which it is possible to change the Rampage processor in a very simple way.
Try and imagine to be able to purchase a modular accelerator on which you can change, whenever you want (or need), GPU/VPU, RAM having as a start base a simple 'raw' video board at a very cheap price to buy… Imagination or future reality?



3DFX Spectre 1000: heat sink and 3DFX Spectre 1000:Rampage chip.

[musmax]

One of the most-advertised feature for Matrox Parhelia offering is its 10-bit per RGBA component processing.

Well, Rampage would have smacked down Matrox since it was set to feature 13-bit per RGBA component, allowing it to have 52-bit internal color processing.
It still made use of 32-bit output, but the quality, and precision of colors would have been of a much higher standard than anything offered by today commercial video card products.

3dfx famous FSAA is implemented as well.
It would have still used RGMS (rotated grid multi sampling), but Rampage texturing abilities would have directly affected how anti-aliasing worked.
The M-Buffer (similar to the Voodoo5 T-Buffer), would have allowed for 4 sample AA per clock with no pixel rate loss, unlike the older T-Buffer, which took a 4x performance hit when AA is enabled.

[Agent500]

Nice Preview, nice card and nice information about the Rampage. Respect!

Niklas

[^SiRtA^]

Originally posted by Agent500
Nice Preview, nice card and nice information about the Rampage. Respect!

Niklas

Thx Agent

[Sansones]

The time to see the result of some test mad has arrived an so sit down and admire this beutiful screenshots...



3DMark 2000 with Direct 3D...





Quake 3 with 1024x768 32bit resolution, trilineare filter and all active...